Assessment Assumptions (See NSCA Text, Essentials . . ., 4th ed., p. 250-52)
Validity (the degree a measure measures what it is supposed to measure)
Reliability (the degree a result is repeatable)
Process (correct methods, techniques, and instruments)
Getting Good Numbers (See NSCA Text, Essentials . . ., 4th ed., p. 253-57)
Assessment is a Negative Sum Process
For a given assessment there exists a best value, a value that is the most accurate representation of a subject's actual measured characteristic. Any variation from perfect adherence in the assessment process or procedure leads to a less accurate representation. You have a maximum potential value and all results move in a negative-accuracy direction depending on your execution of the assessment.
Assessment values are Approximations of True Values (SEE)
Even with perfect assessment implementation, all assessment results are approximate values of a subject's actual measured state. All measures have a range of accuracy defined by the Standard Estimated Error (SEE) of a given assessment(e.g., ± 5 ml · kg-1·min-1). Thus a subject that measures at 45 ml · kg-1·min-1 for a VO2max test could actually have true value that would fall between 40 to 50 ml · kg-1·min-1.
Assessment Statistics
Inferential Statistics (See NSCA Text, Essentials . . ., 4th ed., p. 292-93)
r - Pearson correlation (inter)
relationship of one variable to another variable
r values should not be understood to indicate cause and effect relationships. Confirmation of Cause and effect relationships requires additional experimentation.
values range from -1 to 1
positive r values represent direct relationships in that both variables move in the same directions (i.e., variable 'a' goes up, variable 'b' goes up or variable 'a' goes down, variable 'b' goes down)
negative r values represent inverse relationships (i.e., variable 'a' goes up, variable 'b' goes down or variable 'a' goes down, variable 'b' goes up)
-1 or 1 represents a perfect relationship (i.e., if variable 'a' occurs variable 'b' occurs)
the higher the r value the stronger the relationship between variables.
the lower the r value represents an increasingly random relationship between variables with a r value of zero representing no relationship between variables.
interpretations of 'r'
R - reliability correlation (intra).
relationship between a variable's measure and remeasured overtime.
values range from 1 to 0
value of 1 represents perfect repeatability, perfect reliability (e.g., if you measure variable 'a' today and then measure it the same way tomorrow, it is the same value, always).
the lower the R value, the less reliable and the less repeatable the measure over time (e.g., variable 'a' equals 10 today and tomorrow it measures 15.
lower reliabilities will also have larger SEE's
interpretations of 'R'
p - probability of occurrence
what are the chances that 'r' or 'R' would happen beyond a chance occurrence (it just happened, nothing caused it to happen).
values can range from 1 (typically will not see, I haven't seen it) to very small decimals (.000001) or smaller (p = .05 or p = .002
interpreting 'p': If p = .05 then the relationship indicated by 'r' could occur by chance 5 times out of 100 or stated another way, the relationship indicated occurs 95 times out of 100 beyond chance.
Descriptive Statistics (See NSCA Text, Essentials . . ., 4th ed., p. 291-92)
Central tendency - mean (M), median (Mdn), mode (See NSCA Text, Essentials . . ., 4th ed., p. 291)
Standard Deviation (SD) - measure of variability among values (See NSCA Text, Essentials . . ., 4th ed., pp. 291-92)
Z-score,
T-Scores, Percentage Ranking -- Standardize ways of looking at individuals
data.